This proposal is to expand our understanding of the etiology of celiac disease (CD). CD is unique among autoimmune diseases in that a causative environmental factor is known and its removal is curative, however, other non-genetic factors are likely involved. A major genetic susceptibility locus is known (HLA-DQ), but additional genes are important and remain to be identified. Unfortunately most of the CD patients in the United States remain undiagnosed and untreated. Large groups of high-risk children need to be studied to reveal the full spectrum of the disease, identify its additional genetic and environmental determinants, and develop optimal screening and prevention strategies. We are in a unique position to meet these goals by using population laboratory created by an ongoing large- scale newborn HLA screening funded by the NIDDK (R01, DK-32493, Rewers, M, P.I.). The specific aims of this project are: 1. Determine the prevalence of anti-endomysial antibodies (EMA) in 2600 children at risk for CD by testing: a) at ages 9, 15, 24, 36, and 48 months, 400 children with the HLA-DR3/3 or 314 genotype and a control group of 400 children with the HLA-DR4/x or 414 genotype; b) at ages 24 and 42 months, 1400 children with the HLA-DR3/7, 5/7, or 3/x genotype; c) at ages 9-60 months, an estimated 100 first degree relatives of CD patients and 300 patients with insulin-dependent diabetes mellitus. 2. Perform small bowel biopsy in the expected 70 children positive for EMA, to determine the prevalence of CD. 3. Characterize prospectively (in 800 children) and retrospectively (in 1800 children) dietary factors associated with development of CD, including exposure to dietary provitamins. 4. identify additional (non-HLA-DR/DQ) candidate genetic loci for CD through: a) linkage analysis of a large pedigree with nine CD cases using polymorphic microsatellite markers covering the genome with an average spacing of 11 cM; b) analysis of linkage disequilibrium between candidate CD loci and polymorphic marker loci in an estimated 10C families with a CD proband and at least one of his/her parents available heterozygous for the marker genotype. The proposed research is critical for development of cost-effective population screening for CD which is needed because: 1) CD probably affects 0.2-0.5% of the U.S. population but the clinical diagnosis is often difficult and delayed; 2) CD can cause severe malnutrition, growth impairment and malignancies; 3) early detection is feasible using a sensitive and specific test (EMA); and 4) early treatment is effective. There is a paucity (especially in the U.S) of prospective population studies of dietary and other candidate exposures in children genetically susceptible for CD. Identification of additional environmental and genetic causes of CD will significantly increase prevention opportunities for this as well as other autoimmune diseases.